Surface clear‐sky shortwave radiative closure intercomparisons in the Weather Research and Forecasting model

Abstract

The ability of the Rapid Radiative Transfer Model for GCMs (RRTMG) and Goddard shortwave (SW) radiative schemes to reproduce global (GHI), direct (DNI), and diffuse (DIF) fluxes at the surface in clear‐sky conditions is examined within the Weather Research and Forecasting (WRF) modeling framework. The models were forced using measurements of precipitable water and aerosol optical properties from five AERONET sites and were evaluated against high‐quality observations of GHI, DNI, and DIF at four nearby or coincident sites of the SURFRAD network and one at the Atmospheric Radiation Measurement Central Facility in the contiguous US. In addition, the performance has been intercompared against the Dudhia broadband SW parameterization together with a regressive model to calculate DNI and DIF and the top‐class SMARTS2 SW spectral model. The role of aerosols has been investigated by running the models with and without aerosols. All of them have shown outstanding skill at predicting GHI within the range of the expected observational error, regardless of whether aerosol information is provided as input. When aerosol information is not provided, DNI and DIF are affected by strong biases. In contrast, providing observed aerosols as input solves the bias issue for DNI and nearly does for DIF using the RRTMG scheme. The RRTMG has also proven to have high potential skill for long‐term assessment of clear‐sky GHI, DNI, and DIF irradiance, as long as reliable aerosol inputs are provided. Considering its simplicity, the Dudhia scheme has also shown a remarkable ability for solar resource assessment in clear‐sky conditions.